Re: [Gnash-dev] Re: SIGPIPE on ::writeNet

[strk]

On Wed, Nov 19, 2008 at 07:28:02AM -0700, Rob Savoye wrote:
> strk wrote:
> 
> > Ok, but shouldn't RTMPClient handle it internally rather
> > then letting the signal reach the caller ?
> > It results in an abort here..
> 
>   You shouldn't be getting a signal anyway, as the lower level
> networking code handles the connection closing. As I've never seen this
> behavior when testing against Red5, I'd assume it's related to the other
> changes you've been making.

Well, surely before my changes rtmpget aborted much earlier, so
could not get to that point.

Anyway, client side RTMP libs don't seem to be ready for work in
Gnash, so I can only either join your development efforts in libnet
or wait for you to bring it up to usable state.

--strk;